I try to vote according to policies I think of as “Pro-social Content Promotion” mixed with “Debate Judgement” mixed with “Personal Appreciation Sessions”.
A personal appreciation session involves no downvoting and is the opposite of karmassassination, except its not indiscriminate upvoting—instead I work through someone’s oeuvre looking for unappreciated diamonds to upvote. Mostly I do this for newbies with new and interesting perspectives who I want to come back and say more in the same vein.
The first two policies can involve downvoting. They require making relative judgements (between sibling content for promotion and between parent-child content for judging) and then voting so as to nudge the content closer to my ideal. If a superior comment has few votes relative to a worse comment, I’ll vote the better one up and the worse one down. If I come back later and the relative votes have become too unbalanced some other way (say with sibling comments of mixed quality bubbling into the system), I may reverse votes to bring the total closer to my ideal again. However, I use 0 as a floor, so most of my voting involves trying to pick N/2 comments with 0, 1, or 2 votes to upvote from among the N that I found at that level… to upvote all but one “0 voted” comment would inject relatively few bits into the system, whereas upvoting roughly half of them maximizes the informativeness of my votes :-)
The only things I push negative are things which I believe actively lower the sanity waterline of the community by bringing up issues we are unlikely to be able to handle reasonably (for example “politics is the mind killer” triggers) and then I use −4 as the target value for “the least offensive such thing” and only go deeper into negative if many negative comments are siblings, to express gradations of displeasure (-4, −5, −6… which basically never happens in practice). Mostly I upvote things that are already negative where someone else presumably “didn’t like it” but I think it’s “merely of low quality” rather than actively waterline threatening.
This is exactly what I do, in every detail, except that I use −3 as my least negative target for least offensive, and I have an additional inclination to give a point for effort to what look like extensively edited and-worked on comments that have 0 net votes, even if the obvious effort was not obviously well spent.
I try to vote according to policies I think of as “Pro-social Content Promotion” mixed with “Debate Judgement” mixed with “Personal Appreciation Sessions”.
A personal appreciation session involves no downvoting and is the opposite of karmassassination, except its not indiscriminate upvoting—instead I work through someone’s oeuvre looking for unappreciated diamonds to upvote. Mostly I do this for newbies with new and interesting perspectives who I want to come back and say more in the same vein.
The first two policies can involve downvoting. They require making relative judgements (between sibling content for promotion and between parent-child content for judging) and then voting so as to nudge the content closer to my ideal. If a superior comment has few votes relative to a worse comment, I’ll vote the better one up and the worse one down. If I come back later and the relative votes have become too unbalanced some other way (say with sibling comments of mixed quality bubbling into the system), I may reverse votes to bring the total closer to my ideal again. However, I use 0 as a floor, so most of my voting involves trying to pick N/2 comments with 0, 1, or 2 votes to upvote from among the N that I found at that level… to upvote all but one “0 voted” comment would inject relatively few bits into the system, whereas upvoting roughly half of them maximizes the informativeness of my votes :-)
The only things I push negative are things which I believe actively lower the sanity waterline of the community by bringing up issues we are unlikely to be able to handle reasonably (for example “politics is the mind killer” triggers) and then I use −4 as the target value for “the least offensive such thing” and only go deeper into negative if many negative comments are siblings, to express gradations of displeasure (-4, −5, −6… which basically never happens in practice). Mostly I upvote things that are already negative where someone else presumably “didn’t like it” but I think it’s “merely of low quality” rather than actively waterline threatening.
This is exactly what I do, in every detail, except that I use −3 as my least negative target for least offensive, and I have an additional inclination to give a point for effort to what look like extensively edited and-worked on comments that have 0 net votes, even if the obvious effort was not obviously well spent.